APPLIED PHYSICS LETTERS 90, 121103 共2007兲
Enhanced photoinduced birefringence in polymer-dye complexes:
Hydrogen bonding makes a difference
Arri Priimagia兲 and Matti Kaivola
Department of Engineering Physics and Mathematics and Center for New Materials, Helsinki University
of Technology, P.O. Box 3500, FI-02015 TKK, Finland
Francisco J. Rodriguez and Martti Kauranen
Institute of Physics, Tampere University of Technology, P.O. Box 692, FI-33101 Tampere, Finland
共Received 18 December 2006; accepted 10 February 2007; published online 19 March 2007兲
The authors demonstrate that photoinduced birefringence in azo-dye-doped polymers is strongly
enhanced by hydrogen bonding between the guest molecules and the polymer host. The primary
mechanism behind the enhancement is the possibility to use high dye doping levels compared to
conventional guest-host systems because dye aggregation is restrained by hydrogen bonding.
Moreover, hydrogen bonding reduces the mobility of the guest molecules in the polymer host
leading to a larger fraction of the induced birefringence to be preserved after the excitation light has
been turned off. © 2007 American Institute of Physics. 关DOI: 10.1063/1.2714292兴
Photoinduced optical anisotropy of amorphous
azobenzene-containing polymers has emerged as a source of
exciting optical phenomena,1,2 which are of potential use in
applications ranging from optical data storage3 and
switching4 to waveguiding5 and diffractive optical
elements.6,7 The photoactive properties of such materials
arise from the photoisomerization of the N v N double bond
of the azo molecules. Upon excitation with linearly polarized
light, the molecules with sufficient dipolar component in the
polarization direction undergo successive isomerization
cycles and change their orientation. Due to the selective excitation, the molecules tend to line up perpendicularly to the
polarization of the writing beam, thus becoming insensitive
to the excitation light. The anisotropic molecular orientation
results in stable photoinduced birefringence or dichroism,
which can be optically erased, for instance, by randomizing
the alignment with circularly polarized light.8
In the simplest form of amorphous photoactive polymers, the azo dyes are dissolved in a passive polymer matrix.
Such guest-host systems are flexible and cost effective as
they only require mixing of the constituents to produce a
desired compound. However, their applicability is restricted
by aggregation of the dye molecules due to intermolecular
dipole-dipole interactions, which limits the dye concentration
to moderate levels. The aggregation changes the spectroscopic properties of the material, and also affects the photoinduced reorientation process.9–12 Furthermore, the temporal
and thermal stabilities of the induced anisotropy in guesthost systems are typically poor. The photoalignment and the
subsequent relaxation process is strongly influenced by the
local environment provided by the polymer matrix.13 In particular, by using polymers with high glass-transition temperature 共Tg兲, the stability of the photoinduced orientation
can be increased.14
The drawbacks of guest-host systems have been addressed by covalently bonding the photoactive moieties to
the polymer backbone to form side-chain systems. In such
systems, the dye concentration can be significantly increased,
and efficient photoalignment can be induced even when an
a兲
Electronic mail: [email protected]
azo molecule is attached to each repeating unit of the
polymer.5,9 By combining the high dye concentration with
optimized molecular design, copolymers with exceptionally
large photoinduced birefringence have been synthesized.15–17
Dye aggregation takes place also in side-chain polymers, but
the onset of the dipole-dipole interactions is delayed to
higher concentrations. A further advantage of side-chain
polymers is the long-term stability of the induced anisotropy,
which is significantly higher than in corresponding guesthost systems.14 From a practical point of view, however, covalently linked polymers are considerably less attractive than
guest-host systems, as organic synthesis is required for each
combination of a polymer and an active molecule, which
makes the sample preparation slower and more expensive.
We have previously reported that dye aggregation in a
polymer host can be suppressed by forming hydrogenbonded or protonated polymer-dye complexes.18 Although
weak, spontaneous noncovalent interactions enable high dye
concentrations to be incorporated into the host polymer without sacrificing the ease of processing of conventional guesthost systems. Moreover, Banach et al. have reported that
such specific noncovalent interactions can be used to enhance the electro-optic response of poled guest-host
polymers.19 In this letter, we show that hydrogen bonding
between the dyes and the polymer host and the subsequent
delay in the onset of dye aggregation can be used to obtain
significantly higher photoinduced birefringence values compared to similar guest-host systems where no such interactions occur. Furthermore, the stability of the induced birefringence is enhanced in hydrogen-bonded guest-host
systems, reaching a level comparable to the one traditionally
obtained in covalently linked polymers.
We study the photoalignment of a common azo dye, Disperse Red 1 共DR1兲, doped in different polymers that are
structurally similar but contain different functional groups
共Fig. 1兲. As polymer hosts we used polystyrene 共PS兲, 共M n
= 50 000, Tg ⬇ 100ⴰC兲, poly共4-vinylphenol兲 共PVPh兲, 共M n
= 1100– 5200, Tg ⬇ 120°C兲, and poly共4-vinylpyridine兲
共P4VP兲 共M n = 50 000, Tg ⬇ 140°C兲. PS is a nonpolar polymer
that acts as an inert reference with no significant interactions
with the polar DR1 molecules. PVPh and P4VP are polar
0003-6951/2007/90共12兲/121103/3/$23.00
90, 121103-1
© 2007 American Institute of Physics
Downloaded 30 Mar 2007 to 130.230.131.201. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
121103-2
Appl. Phys. Lett. 90, 121103 共2007兲
Priimagi et al.
FIG. 1. Chemical structures of the materials used.
polymers containing functional groups that enable specific
intermolecular interactions between the polymer host and the
DR1 molecules. More precisely, the phenol and pyridine
groups of PVPh and P4VP can form hydrogen bonds with
the nitrobenzene and hydroxyl groups of DR1,
respectively.19,20
Thin films with different doping levels of DR1 in the
polymer hosts were prepared by spin coating the constituents
from dimethylformamide 共DR1 in PVPh/P4VP兲 or tetrahydrofuran 共DR1 in PS兲 onto clean glass substrates. The film
thicknesses were measured with a DEKTAK 3 surface profiler, and they ranged from 70 to 700 nm. The thicknesses
were varied from sample to sample to maintain constant optical density at the writing wavelength 共532 nm兲, which is
necessary for meaningful comparison of the photoinduced
anisotropy in different samples.21 UV-vis absorption
spectra were taken with a Perkin-Elmer Lambda 950
spectrophotometer.
The photoalignment was performed by using an
s-polarized 共normal to the plane of incidence兲 beam from a
continuous-wave Nd:YVO laser 共532 nm兲 with an angle of
incidence of approximately 10°. The process was monitored
with a diode laser 共850 nm兲, normally incident on the
sample. The transmitted intensity of the monitoring beam
through a polarizer-sample-analyzer combination was measured with a photodiode to probe the photoinduced anisotropy. The orientation of the polarizer/analyzer was set to
±45° with respect to the polarization direction of the writing
beam. The birefringence 兩⌬n兩 can be obtained from the transmission data as
I = I0 sin2
冉
冊
␲兩⌬n兩d
,
␭
共1兲
where I0 is the photodiode signal for parallel polarizer/
analyzer orientation 共in the absence of the sample兲, d is the
film thickness, and ␭ is the wavelength of the monitoring
beam.
The absorption spectra of the samples containing different concentrations 共5 – 30 wt % 兲 of DR1 doped in PS, PVPh,
FIG. 3. Photoinduced birefringence as a function of dye concentration for
DR1 in PS, P4VP, and PVPh. The reported values are the saturation values
of the birefringence while the writing beam is applied on the sample. The
writing beam intensity was 150 mW/ cm2.
and P4VP are presented in Fig. 2. In PS, DR1 aggregation
results in a significant blueshift of the absorption maximum
at concentrations exceeding 10 wt %. In PVPh and P4VP
this behavior is absent, indicating that most of the dye molecules remain isolated in the investigated concentration
range. The reduced aggregation in the active host polymers is
attributed to the coupling of DR1 to the polymer chains
through specific noncovalent interactions, providing a more
favorable environment for the dye molecules.18,19
Figure 3 shows the photoinduced birefringence of DR1
in PS, PVPh, and P4VP as a function of dye concentration.
At 5 wt % DR1 concentration, the birefringence is essentially the same in each matrix. However, at higher concentrations the difference between the polymers is significant. In
PS the maximum birefringence 共aproximately 0.01兲 is
achieved at 10– 20 wt % after which it decreases to 0.005 at
30 wt %. In PVPh birefringence increases approximately linearly, reaching a value ⬎0.04 at the 30 wt % DR1 concentration. In P4VP the behavior is similar, aside from the slight
saturation observed at high concentrations. This result correlates well with the absorption spectra of Fig. 2 and addresses
the destructive impact of dye aggregation on the photoinduced anisotropy. We also remark that the 30 wt % DR1 concentration corresponds approximately to a molar ratio of
0.15:1 of DR1:polymer unit. We have shown earlier that by
exploiting stronger noncovalent interactions such as proton
transfer, the onset of aggregation can be delayed until essentially all functional groups of the polymer are occupied 共1:1
molar ratio兲.18 Thus we expect that by optimizing the functional groups of the material system, the range of linear
growth of the birefringence can be extended to even higher
concentrations.
Apart from providing the possibility to use higher dye
concentrations, specific noncovalent interactions can de-
FIG. 2. Absorption spectra of thin films of 共a兲 DR1 in PS, 共b兲 DR1 in PVPh, and 共c兲 DR1 in P4VP.
Downloaded 30 Mar 2007 to 130.230.131.201. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
121103-3
Appl. Phys. Lett. 90, 121103 共2007兲
Priimagi et al.
stability is comparable to the stability of similar side-chain
polymers—more than 80% of the birefringence is conserved
after the excitation beam is turned off. The concept provides
a particularly simple method to obtain high and stable photoinduced birefringence by using guest-host polymers and is
potentially useful in various applications where dye aggregation can be a limiting factor for the system performance.
FIG. 4. Writing/relaxation curves for samples containing 5 wt % of DR1 in
PS, PVPh, and P4VP. The irradiation period was 0 – 1500 s and the writing
beam intensity was 30 mW/ cm2.
This work was partially supported by Academy of Finland 共108538兲. One of the authors 共A.P.兲 acknowledges the
financial support of the Finnish Cultural Foundation. The
authors also thank B. Löfgren, K. Lindfors, A. Laiho, and A.
Shevchenko for comments and discussion and M. Annala for
technical assistance.
1
crease the mobility of the dyes, thus improving the temporal
stability of their net alignment.19 In our case, dye concentration played no essential role in relaxation, implying that plasticization effects due to addition of the dye are insignificant.
Hence, we only consider the samples containing 5 wt % of
DR1 共each having essentially the same birefringence兲 in
more detail. Figure 4 shows the normalized writing/
relaxation curves for these samples. The writing process is
seen to be slower in PVPh and P4VP than in PS, which could
be due to the lower mobility of the DR1 molecules in the
interacting polymers. However, our preliminary results on
other polymer-dye complexes suggest that this limitation
could be overcome simply by choosing a proper pair of materials. The most notable difference between the polymers is
in the stability of the birefringence. The fitting of the normalized decay curves with a biexponential function yields a remnant birefringence of 64% from the saturation value in PS,
while in PVPh and P4VP this value exceeds 80%. Such a
high stability is comparable to the values of DR1-containing
side-chain polymers.9 This result emphasizes that spontaneous noncovalent interactions provide a promising alternative
to the more traditional approach based on covalent interactions. Note also that the difference in the long-term stability
probably cannot be accounted for by the small differences in
the glass-transition temperatures; for PS and PVPh this difference is only approximately 20 ° C.
In conclusion, we have shown that hydrogen bonding
between azo dyes and polymer chains can be used to enhance
the photoinduced anisotropy in guest-host polymers. Such
specific interactions allow the dye concentration to be increased without aggregation, resulting in birefringence value
of 0.04 at 30 wt % DR1 in PVPh. Moreover, the long-term
Photoreactive Organic Thin Films, edited by Z. Sekkat and W. Knoll
共Academic, San Diego, 2002兲.
2
A. Natansohn and P. Rochon, Chem. Rev. 共Washington, D.C.兲 102, 4139
共2002兲.
3
S. J. Zilker, T. Bieringer, D. Haarer, R. S. Stein, J. W. van Egmond, and S.
G. Kostromine, Adv. Mater. 共Weinheim, Ger.兲 10, 855 共1998兲.
4
Y. Luo, W. She, S. Wu, F. Zeng, and S. Yao, Appl. Phys. B: Lasers Opt.
80, 77 共2005兲.
5
Y. Shi, W. H. Steier, L. Yu, M. Chen, and L. R. Dalton, Appl. Phys. Lett.
59, 2935 共1991兲.
6
A. Natansohn and P. Rochon, Adv. Mater. 共Weinheim, Ger.兲 11, 1387
共1999兲.
7
R. J. Stockermans and P. L. Rochon, Appl. Opt. 38, 3714 共1999兲.
8
A. Natansohn, P. Rochon, J. Gosselin, and S. Xie, Macromolecules 25,
2268 共1992兲.
9
D. Brown, A. Natansohn, and P. Rochon, Macromolecules 28, 6116
共1995兲.
10
F. L. Labarthet, S. Freiberg, C. Pellerin, M. Pézolet, A. Natansohn, and P.
Rochon, Macromolecules 33, 6815 共2000兲.
11
I. Zebger, M. Ruloth, U. Hoffmann, J. Stumpe, H. W. Seisler, and S.
Hvilsted, J. Phys. Chem. A 106, 3454 共2002兲.
12
D. S. Dos Santos, Jr., C. R. Mendonça, D. T. Balogh, A. Dhanabalan, J. A.
Giacometti, S. C. Zilio, and O. N. Oliveira, Jr., Polymer 43, 4385 共2002兲.
13
K. Tawa, K. Kamada, T. Sakaguchi, and K. Ohta, Polymer 41, 3235
共2000兲.
14
A. Natansohn, P. Rochon, C. Barrett, and A. Hay, Chem. Mater. 7, 1612
共1995兲.
15
T. Fukuda, J. Y. Kim, D. Barada, T. Senzaki, and K. Yase, J. Photochem.
Photobiol., A 182, 262 共2006兲.
16
J. Y. Kim and T. Fukuda, Mol. Cryst. Liq. Cryst. 446, 71 共2006兲.
17
B. L. Lachut, S. A. Maier, H. A. Atwater, M. J. A. de Dood, A. Polman,
R. Hagen, and S. Kostromine, Adv. Mater. 共Weinheim, Ger.兲 16, 1746
共2004兲.
18
A. Priimagi, S. Cattaneo, R. H. A. Ras, S. Valkama, O. Ikkala, and M.
Kauranen, Chem. Mater. 17, 5798 共2005兲.
19
M. J. Banach, M. D. Alexander, Jr., S. Caracci, and R. A. Vaia, Chem.
Mater. 11, 2554 共1999兲.
20
W. F. Baitinger, P. von R. Schleyer, T. S. S. R. Murty, and L. Robinson,
Tetrahedron 20, 1635 共1964兲.
21
P. Rochon, D. Bissonnette, A. Natansohn, and S. Xie, Appl. Opt. 32, 7277
共1993兲.
Downloaded 30 Mar 2007 to 130.230.131.201. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp